Practical considerations in choosing between the case-cohort and nested case-control designs.
Explore the source record for details and available documents.
Case-Control Studies↗
Biomedical subjects
S Wacholder
Publications and source records attributed to S Wacholder.
Selecting an efficient design for assessing exposure-disease relationships in an assembled cohort.
We develop approximate methods to compare the efficiencies and to compute the power of alternative potential designs for sampling from a cohort before beginning to collect exposure data. Our methods require only that the cohort be assembled, meaning that the numbers of individuals Nkj at risk at pairs of event times tk and tj greater than or equal to tk are available. To compute Nkj, one needs to know the entry, follow-up, censoring, and event history, but not the exposure, for each individual. Our methods apply to any "unbiased control sampling design," in which cases are compared to a random sample of noncases at risk at the time of an event. We apply our methods to approximate the efficiencies of the nested case-control design, the case-cohort design, and an augmented case-cohort design, compared to the full cohort design, in an assembled cohort of 17,633 members of an insurance cooperative who were followed for mortality from prostatic cancer. The assumptions underlying the approximation are that exposure is unrelated both to the hazard of an event and to the hazard for censoring. The approximations performed well in simulations when both assumptions held and when the exposure was moderately related to censoring.
Biometry↗
Diet, tobacco use, and fatal prostate cancer: results from the Lutheran Brotherhood Cohort Study.
A cohort of 17,633 white males age 35 and older responded to a mailed epidemiological questionnaire in 1966 and was followed until 1986 to determine the risk of cancer associated with diet, tobacco use, and other factors. During the 20-year follow-up, 149 fatal prostate cancer cases were identified. Relative risks for prostate cancer were significantly elevated among cigarette smokers (relative risk, 1.8; 95% confidence interval, 1.1-2.9) and users of smokeless tobacco (relative risk, 2.1; 95% confidence interval, 1.1-4.1). No significant associations were found with frequency of consumption of meats, dairy products, fruits, or vegetables. There were no overall significant associations between consumption of vitamin A from animal sources (retinol) and provitamin A from plant sources (carotene) and risk, but positive trends were seen for ages under 75, while inverse associations were found at older ages. Beverage consumption, including drinking coffee and alcohol, was unrelated to risk. Marital status, education, rural/urban status, and farming residence were also unrelated to the risk of fatal prostate cancer. The findings add to limited evidence that tobacco may be a risk factor for prostate cancer, but fail to provide clues to dietary or other risk factors.
Adult↗
Does nondifferential misclassification of exposure always bias a true effect toward the null value?
The authors present some examples to demonstrate that in certain nondifferential misclassification conditions with polychotomous exposure variables, estimates of odds ratios for categories at intermediate level of risk can be biased away from the null or can change direction. In addition, the authors present two examples to demonstrate that the slope of the dose-response trend for the true distributions can change direction, creating a false inverse trend, even if the misclassification is nondifferential.
Bias↗
Re: "Case-control studies using other diseases as controls: problems of excluding exposure-related diseases".
Explore the source record for details and available documents.
Case-Control Studies↗
The design and analysis of case-control studies with biased sampling.
A design is proposed for case-control studies in which selection of subjects for full variable ascertainment is based jointly on disease status and on easily obtained "screening" variables that may be related to the disease. Recruitment of subjects follows an independent Bernoulli sampling scheme, with recruitment probabilities set by the investigator in advance. In particular, the sampling can be set up to achieve, on average, frequency matching, provided prior estimates of the disease rates or odds ratios associated with screening variables such as age and sex are available. Alternatively--for example, when studying a rare exposure--one can enrich the sample with certain categories of subject. Following such a design, there are two valid approaches to logistic regression analysis, both of which allow for efficient estimation of effects associated with the screening variables that were allowed to bias the recruitment. The statistical properties of the estimators are compared, both for large samples, based on asymptotics, and for small samples, based on simulations.
Case-Control Studies↗
Dual responses to increase validity in case-control studies.
This paper compares three methods to increase validity of case-control studies when two imperfect measurement tools are available to assess exposure. From simulated data, we show that point estimates of the relative risk, using only concordant dual responses, are always less biased than any other estimate. Because of the loss of power due to discarding discordant responses, the concordant response design is not efficient and is expensive. We show that a series response estimation, whereby all information is used but only dual positive responses are considered exposed, can be used to complement the interpretation of the less biased estimates. This approach is more interesting when exposure is rare and measuring it expensive.
Bias↗
A case-control study of cancer of the esophagus and gastric cardia in Linxian.
A case-control study involving interviews with 1,244 patients (758 males and 486 females) with cancer of the esophagus or gastric cardia and 1,314 population-based controls (789 males, 525 females) was carried out in Linxian, a rural county in North Central China with one of the world's highest mortality rates for these tumors. Cancer risks tended to rise with increasing intake of wheat and corn, but no association was found with adult intake of pickled vegetables, the leading a priori suspect, and risks were not elevated among those consuming low quantities of fresh vegetables or fruits. Few differences in preparation or storage of food or water were detected, although cancer patients reported less fluid intake than controls. Few persons reported drinking alcoholic beverages. Smoking was reported by 61% of the male cases and was a mild risk factor, related more to cancer of the cardia than of the esophagus. The risk was increased by 70% among those whose parents had esophageal or stomach cancer, but only slightly among those whose spouses had such cancers, suggesting that exposure early in life and/or genetic effects may be involved.
Adult↗
Fecapentaene concentration and mutagenicity in 718 North American stool samples.
The fecapentaenes (FP) are the predominant fecal mutagens identified to date, but they have not been shown to be carcinogenic. Epidemiologists looking for other fecal mutagens that may be related to colorectal cancer must disentangle from their investigations the pervasive mutagenic effect of the fecapentaenes. As a first step to studying the epidemiology of fecal mutagenicity independent of fecapentaenes, we compared FP measurements and Salmonella mutagenicity assay results for 718 acetone-extracted stool samples collected from a variety of subjects in the Washington DC metropolitan areas. In this large group, 50% of mutagenic samples contained elevated fecapentaenes. Specifically, three-quarters of the samples mutagenic in TA100 contained high FP levels. In contrast, mutagenicity in TA98 was not generally explainable by fecapentaenes, suggesting that non-fecapentaene TA98 mutagenicity should be one focus of future efforts to uncover colorectal carcinogens of etiologic importance.
Feces↗
Smoking and degree of occupational exposure: are internal analyses in cohort studies likely to be confounded by smoking status?
Occupational cohort studies are usually carried out without the benefit of information on smoking habits of cohort members. One common approach to avoid confounding bias related to smoking habits is to carry out an internal analysis, comparing workers with different degrees of occupational exposure. The premise behind this approach is that within a cohort there is unlikely to be correlation between degree of exposure and smoking habits. If this were untrue, smoking could confound the disease-exposure relationships. Our purpose was to verify the premise. The study sample consisted of 857 French-Canadian men born between 1910 and 1930, with 11 or fewer years of education, and interviewed around 1980 in the context of an occupational cancer case-control study. For each man we had information on smoking habits, job history, and a history of the chemicals he was exposed to in each of his jobs. We computed two indices of the dirtiness of workers' job histories: one based on the job titles held by the man and a second based on the degree of exposures to workplace substances. There was no correlation between these indices of job dirtiness and smoking history. We also examined the smoking-exposure relationship among the subsets of men who had been occupationally exposed to ten especially noticeable substances. Within the subsets, there was no indication of a consistent difference among the smoking subgroups in level or duration of exposure to these index substances. These findings do not support the view that nonsmokers sought out cleaner job environments than smokers; they imply that internal analyses of "dose-response" in cohort studies are unlikely to be seriously confounded by smoking habits.
Adult↗
Indirect corrections for confounding under multiplicative and additive risk models.
We define a multiplicative model and an additive model for the hazards associated jointly with exposure and with the presence of a confounder like smoking. Under the multiplicative model, the crude relative risk may be adjusted indirectly, by means of a factor proposed by Axelson [1978], and implicitly by Cornfield et al. [1959] and Schlesselman [1978]. We present corresponding indirect correction formulas under the additive risk model for the risk difference and for the excess relative risk. Conditions are established under which these corrections may be applied to age-adjusted rates from composite study populations. We demonstrate that indirect corrections may be no better than crude measures of risk if one assumes the wrong model for the joint action of the exposure and confounding factors. These results are illustrated on an example of occupational exposure to vermiculite. The limitations of the techniques are discussed.
Epidemiologic Methods↗
Degree of confounding bias related to smoking, ethnic group, and socioeconomic status in estimates of the associations between occupation and cancer.
In occupational cancer epidemiology, many studies are carried out without access to information on smoking and other potential confounding variables. It is unclear whether such deficiencies are likely to cause serious bias in estimates of cancer-occupation associations. An empiric investigation was carried out to determine the effect of inclusion or exclusion of three variables--smoking, ethnic group, and socioeconomic status--on estimates of odds ratios (OR) between 25 occupations and three types of cancer--lung, bladder, and stomach. Of the 75 associations studied, only one OR was distorted by more than 40% when comparing unadjusted with adjusted estimates; three were distorted by between 30% and 40%; four others by between 20% and 30%. Of the eight associations which were distorted by more than 20%, seven involved lung cancer and one involved bladder cancer; none involved stomach cancer. An additional analysis was carried out on the 25 lung cancer-occupation associations to determine whether the nature of the stratification on smoking (ie, whether crude or "precise" categories were used) gave different OR estimates. The differences in ORs induced by different parametrizations of the smoking variable were relatively small. Our results support the view that relative risks between lung cancer and occupation in excess of 1.4 are unlikely to be artifacts due to uncontrolled confounding. For bladder and stomach cancer, the corresponding cut point may be as low as 1.2. In studies of occupation and cancer, uncontrolled confounding due to smoking and social class may not be as serious a threat to the integrity of results as is sometimes feared.
Adult↗
Visits to physicians before and after exposure to urea formaldehyde foam insulation.
The average number of visits to a physician made by a sample of 351 residents of homes insulated with urea formaldehyde foam insulation in Montreal in the one year period before exposure was 5.25, and in the year following 5.62, an increase of 7 per cent (odds ratio 1.07, 95% CI = 1.00, 1.15). The increase in visits in the post insulation year was limited to subjects who had the product installed in the winter (OR = 1.48, 95% CI = 1.18,1.85), and was not seen for study subjects who insulated their homes during other seasons of the year.
Construction Materials↗
[Determination of the risk of a 2d cancer in patients treated for a first cancer].
A paradoxical effect of radiotherapy and chemotherapy for cancer is that some of these treatments can themselves cause new cancers. Most epidemiologic methods can be applied successfully to the investigation of this problem and this paper reviews various approaches that have already been used by various researchers. The authors first review the more traditional methods, i.e., cohort and case-control studies and they then describe designs that have been proposed more recently, such as case-cohort studies. A distinction is established between internal comparisons, carried out within the study population, and external comparisons, in which a general population external to the population under study is used as the reference category. This presentation is mainly aimed at investigators using tumor registry data. However, the general principles formulated here are easily generalized to contexts other than that of registries.
Antineoplastic Agents↗
Maternal age and primary cesarean section rates: a multivariate analysis.
To study the effect of maternal age on primary cesarean section rates, 3458 consecutive deliveries were analyzed. The cesarean section rates in primiparous women less than 25, 25 to 34, and over 34 years of age were 13.1%, 18.5%, and 28.2%, respectively. A similarly dramatic rise with advancing maternal age was seen in multiparous women with rates of 3.4%, 4.7%, and 10.1%, respectively, in the three age groups. The strong association between cesarean section and maternal age persisted after multivariate adjustment for induction of labor, epidural anesthesia, meconium-stained amniotic fluid, and fetal distress, and thus these factors do not explain the relationship.
Abruptio Placentae↗
External comparisons with the case-cohort design.
The case-cohort design can be an economic alternative to the standard cohort design. Prentice (Biometrika 1986;73:1-11) showed how the case-cohort design can be used to obtain relative risk estimates for comparisons within the cohort being studied. In this paper, the authors consider ways in which the case-cohort design can be used for comparing risk in exposure groups within the cohort to the risk in an external population. The problem reduces to estimating the number of expected cases at each exposure level in the total cohort, when exposure status is available only for members of a subcohort, i.e., a random sample of the total cohort. The authors describe theoretical and empirical properties of several variations of the design and analysis of case-cohort studies. Empirical properties were examined by replicating the selection of the subcohort in a study of second cancer risk after chemotherapy for a first cancer. Use of a case-cohort design in that study would have saved five-sixths of the cost of gathering covariate information at the price of only an 11% loss in efficiency relative to a full cohort study.
Adult↗
Associations between several sites of cancer and twelve petroleum-derived liquids. Results from a case-referent study in Montreal.
A population-based case-referent study provided information on the associations between several types of cancer and 12 petroleum-derived liquids. All site-exposure combinations were investigated. The most interesting results concerned the following combinations: leaded gasoline-stomach cancer, aviation gasoline-kidney cancer (and the possible implications of this association for a similar effect of unleaded automotive gasoline), mineral spirits-squamous-cell cancer of the lung, diesel fuel-nonadenocarcinoma lung cancer, lubricating oils-squamous-cell lung cancer, cutting fluids-bladder cancer, other mineral oils-bladder cancer, mineral spirits-prostate cancer, diesel fuel-prostate cancer, and lubricating oils-prostate cancer.
Humans↗
Discovering carcinogens in the occupational environment. Methods of data collection and analysis of a large case-referent monitoring system.
A multi-cancer site, multi-factor, case-referent study was undertaken to generate hypotheses about possible occupational carcinogens. About 20 types of cancer were included. Incident cases among men aged 35-70 years and diagnosed in any of the major Montreal hospitals were eligible. Probing interviews were carried out for 3 726 eligible cases. The interview was designed to obtain detailed lifetime job histories and information on potential confounders. Each job history was reviewed by a team of chemists who translated it into a history of occupational exposures. These occupational exposures were then analyzed as potential risk factors in relation to the sites of cancer included. For each site of cancer analyzed, referents were selected from among the other sites in the study. The analysis was carried out in stages. First a Mantel-Haenszel analysis was undertaken of all cancer-substance associations, stratifying on a limited number of covariates, and, then, for those associations which were noteworthy in the initial analysis, a logistic regression analysis was made taking into account all potential confounders. This report describes the fieldwork and analytical methods.
Carcinogens, Environmental↗